mstar.model.omnivoice.components

Modules

backbone	The OmniVoice backbone, run as one packed bidirectional forward per step.
codec	Higgs-Audio-v2 codec wrappers: waveform <-> 8-row token canvas.
text	Prompt construction: style tokens, text tokens, and the canvas prefix.
unmask	The unmask step: CFG scoring, confidence ranking, and the reveal schedule.